You are on page 1of 14

This article was downloaded by: [McGill University Library]

On: 12 December 2012, At: 10:57

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office:
Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Construction Management and Economics

Publication details, including instructions for authors and subscription

Selecting a suitable procurement method for

a building project
P.E.D. Love, M. Skitmore & G. Earl
Version of record first published: 21 Oct 2010.

To cite this article: P.E.D. Love, M. Skitmore & G. Earl (1998): Selecting a suitable procurement method for
a building project, Construction Management and Economics, 16:2, 221-233

To link to this article:


Full terms and conditions of use:

This article may be used for research, teaching, and private study purposes. Any substantial
or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or
distribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that
the contents will be complete or accurate or up to date. The accuracy of any instructions,
formulae, and drug doses should be independently verified with primary sources. The publisher
shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages
whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of
the use of this material.
Construction Management and Economics (1998) 16, 221± 233

Selecting a suitable procurement method for a

building project
P.E.D. LOVE 1 , M. SKITMORE 2 and G. EARL 2
1Schoolof Architecture and Building, Deakin University, Geelong 3217, Australia
School of Construction Management, Queensland University of Technology, Brisbane, Queensland 4000, Australia

Received 13 February 1996; accepted 12 June 1997

Downloaded by [McGill University Library] at 10:57 12 December 2012

Building procurement has become a fashionable term with industry practitioners and researchers. It deter-
mines the overall framework and structure of responsibilities and authorities for participants within the
building process. It is a key factor contributing to overall client satisfaction and project success. The selec-
tion of the most suitable procurement method consequently is critical for both clients and project partici-
pants, and is becoming an important and contemporary issue within the building industry. The problem,
nevertheless, lies in the fact that there has been limited empirical research in this ® eld of study. Postal ques-
tionnaire surveys of 41 clients and 35 consultants were carried out, and were used to obtain experience
of and attitudes to a variety of procurement methods and the criteria used for selection. The ® ndings indi-
cate that a simple set of the criteria generally is adequate and suf® cient for procurement path selection, and
that there is a reasonable consensus on the appropriate weighting for each path. Moreover, it is shown that,
contrary to expectations, similar clients generally do not have similar procurement needs.

Keywords: Procurement selection, criteria weighting, client needs, utility rating

Introduction ority rating for project attributes such as speed, cer-

tainty, ¯ exibility, quality, complexity, risk avoidance
A project may be regarded as successful if the building and responsibility, price competition and disputes/arbi-
is delivered at the right time, at the appropriate price and tration. For this procedure to be of practical use it is nec-
quality standards, and provides the client with a high essary ® rst to ® x the weighting factors that relate these
level of satisfaction (e.g. Barclay, 1994). One important attributes to individual procurement methods indepen-
in¯ uence on this, identi® ed in the Banwell and Emerson dent of individual projects.
reports of the 1960s, is the type of procurement method One problem with this, however, is that the factor
implemented. These have proliferated in recent years weights cannot be obtained easily by objective means
and their characteristics have become a major ® eld of and must to elicited from practitioners in the ® eld,
study in their own right (e.g. Nahapiet and Nahapiet, who have found dif® culty in reaching a consensus on
1985; Franks, 1990; Turner, 1990). One result of this is such matters (Hamilton, 1987). A further problem is
a consensus that there is one procurement method that that the client priority ratings have to be established for
is in some sense `better’ than all others for an individual each project. This can be exacerbated further for clients
project but that no one procurement method is likely to who may not have the necessary experience even to pro-
be better than others for any project. Several studies duce an adequate brief. Nahapiet and Nahapiet (1985),
have considered how this `best’ individual procurement however, found the main factors affecting choice of
method may be identi® ed (e.g. NEDO, 1983) by procurement method to be the characteristics of the
reference to a set of project characteristics, attributes or client together with the project characteristics and
criteria. The most advanced of these are those of requirements, suggesting that similar clients with
Skitmore and Marsden (1988) and Singh (1990), who similar project requirements may have similar and con-
propose a procedure involving weighting factors and pri- sistent priority ratings.
0144± 6193 € 1998 E & FN Spon
222 Love et al.

The purpose of the study described in this paper

was to address these two problems in the context of
Australian, and speci® cally the state of Queensland,
practice. A postal questionnaire survey of 35 consul-
tants was conducted to establish if a reasonable con-
sensus existed on factor weightings, while a similar
postal questionnaire survey was conducted with a sam-
ple of 41 clients to establish if a reasonable consensus
existed amongst subgroups of clients on priority ratings
for similar types of projects. Attitudes and experiences
to a variety of procurement methods and criteria used
for selection were derived from the questionnaires. The
results of the research indicate that, for the samples
involved, a reasonable consensus did exist for the factor
weightings but not for priority ratings.
Downloaded by [McGill University Library] at 10:57 12 December 2012

Figure 1 Categorization of building procurement systems

(adapted from Perry (1985))
Building procurement systems

A common leitmotiv of the construction industry is the budget and to be of the highest quality. However, some
proliferation of de® nitions of a procurement system. clients stress that certain criteria are more important
The terms `contractual arrangement’ and `procurement than others. There are numerous derivatives to each
system’ are usually used synonymously. The de® nition procurement method. However, those which have been
adopted here is that a procurement system is `an orga- categorized are considered to be the most popular
nizational system that assigns speci® c responsibil- methods at the time of evaluation.
ities and authorities to people and organizations, and
de® nes the relationships of the various elements in the
Selection of procurement methods in Australia
construction of a project’ .
The most commonly used procurement methods in
(1) Procurement systems can be categorized as:
Australia, based on Ireland (1982) and Barclay (1994)
(2) traditional (design± tender± construct) methods:
design and construct methods; or
(3) management methods. (1) single lump sum contracts and full documenta-
Sub-classi® cations of these systems proliferate within
(2) provisional or partial quantities;
the Australian industry (Figure 1). Novation and
(3) cost reimbursement;
design and manage methods are some examples. It is
(4) package deals/turnkey;
common for procurement systems, contract forms and
(5) construction management; and
price determination mechanisms to be regarded as
(6) management contracting.
synonymous or inextricably related (Fellows, 1993;
Hibberd and Basden, 1996). Project management is excluded as it is considered that
Procurement systems have become increasingly a project manager could be applied to any procurement
¯ exible. Fellows (1993) suggests that the interchange method. In other words, to dispel a common miscon-
that exists between such systems has made it essential ception, project management is not a procurement
to distinguish the procurement system from the formal method (Bennett, 1986, p. 5). The term merely means
subsystem. It is suggested further by Fellows that the that the client has employed an agent to assist in under-
subsystem may be used interchangeably to enable taking a supervisory and coordination role within the
the procurement system to be tuned to the clients’ project. To the above list can be added novation, design
circumstances and requirements. A primary issue that and manage, and contractors design and build.
often is raised within the construction industry relates
to what clients want in order to be satis® ed with their
buildings and the means by which those buildings Selection of criteria
have been procured. Consequently, it is important to
evaluate the clients’ criteria, their importance and then The following criteria can be used to examine
seek performance to match the criteria. All clients client requirements and `experts’ preferences for the
require their buildings to be completed on time, within performance of each procurement method. NEDO
Procurement selection 223

(1985), Skitmore and Marsden (1988) and Singh seeking advice from a number of one or more pro-
(1990) suggest employing the following criteria to fessions. Clients of every kind need to recognize the
establish a pro® le of the clients’ requirements: positive and sustained contribution they have to make
if buildings of excellence are to be the norm rather
(1) speed (during both design and construction;
than the exception. Typically, clients who build on a
(2) certainty (price and the stipulated time and
regular basis may use a system that is compatible
knowledge of how much the client has to pay
with their corporate environment and from advice
at each period during the construction phase);
given from external consultants. Moreover, experi-
(3) ¯ exibility in accommodating design changes;
enced clients may establish a parochial approach to
(4) quality (contractors’ reputation, aesthetics and
building, utilizing those procurement methods that are
con® dence in design);
most familiar. Bresnen and Haslam (1991) de® ne such
(5) complexity (client may specify particular
an approach as `habituation’ , suggesting that this is
subcontractor, or constructability analysis);
inappropriate in situations where a different building
(6) risk allocation/avoidance;
type is required. Whatever decision a client makes
(7) responsibility (completion of program, price,
pertaining to their procurement selection it will have
product quality, design and construction);
Downloaded by [McGill University Library] at 10:57 12 December 2012

a signi® cant effect on the project team and the ¯ ow of

(8) price competition (covering such issues as value
communication between project members.
for money, maintenance costs and competitive
Communication has been linked to team effective-
tendering); and
ness, the integration of work units across organiza-
(9) disputes and arbitration.
tional levels, characteristics of effective supervision,
The use of multiple criteria to derive a suitable pro- job satisfaction, and overall organizational effectiveness
curement method for a building project will assist the (Shockley-Zalaback, 1991). In fact, the project orga-
client in identifying their principal goals and objectives. nization consists essentially of complex communication
processes that create and change events. Consequently,
improved organizational communication will play
Alternative approaches to procurement a signi® cant part in determining the effectiveness of a
selection building project and improving the likelihood of client
satisfaction, especially if the client is involved actively
The approaches for procurement selection range from with the project’ s delivery processes (Masterman and
simple (Franks, 1990) to highly complex (Skitmore and Gameson, 1994; Walker, 1994a; Love and Mohamed,
Marsden, 1988; Liu, 1994). However, it is important 1996). Similarly, Graves (1978) suggests that the stan-
that method selection is done logically, systematically dard of service offered by the building industry relates
and in a disciplined manner by the client’ s principal directly to the amount of effort expended by the client
adviser. Numerous techniques exist. The range in in establishing a good brief, and that satisfaction at the
choice of a procurement system is now so wide and construction stage is linked closely to the degree of
projects are becoming so complex that the selection control and supervision by the clients themselves.
process needs to be carried out in a disciplined and In an attempt to derive simple approaches to select-
objective manner within the framework of the clients’ ing a procurement method for clients, researchers
overall strategic project objectives. Nevertheless, the sometimes have ignored the dif® culties highlighted
major dif® culties associated with procurement selection above. NEDO (1975) suggests that to ensure a suc-
include: (a) no single person, or knowledge `czar’ , has cessful choice of building procurement method, the
been found who is familiar with all the primary pro- client’ s brief must be clear and comprehensive in spec-
curement methods (Hamilton, 1987); (b) no consensus ifying their speci® c needs and objectives. Nevertheless,
has been found between `experts’ which easily system- identifying the client’ s most important needs and
izes procurement selection; and (c) no mutually exclu- objectives can be a problematic task. Some needs
sive sets of criteria uniquely and completely determine and objectives will have to be sacri® ced in preference to
the appropriate procurement method for a speci® c others to suit their corporation’ s particular commercial
project (Ireland, 1985). interests and technical requirements (Walker, 1989).
Despite the dif® culties associated with selecting NEDO (1985) relates the characteristics of the most
a procurement method, Masterman and Gameson popular procurement methods used to a list of nine
(1994) suggest the main in¯ uencing factor of procure- client priorities or needs. This technique, while useful
ment selection is determined by the level of client as a guide in terms of eliminating unsuitable procure-
experience. There is no de® nition as such for the ment methods from the available alternatives, is insuf-
client. John Brandonburger (cited in Blackmore, 1990) ® ciently sophisticated to enable a ® nal decision to be
describes the client as an assorted collection of persons taken as to the method appropriate for a building
224 Love et al.

project (Masterman, 1992). Franks (1990) uses a Love (1996) found that a well established and
rating system based on the ability of each procurement prominent Australian project management organiza-
system to meet seven common satisfying criteria. A tion selected procurement methods for their clients
scale of 1± 5 is used, where 1 is the minimum and 5 using a systematic ® rst-principle analysis, by:
is the maximum. Masterman (1992) states that the use
(1) de® ning the project;
of this technique in determining clients’ needs is valid,
(2) determining the project needs;
but is ¯ awed with subjectivity.
(3) establishing a program;
Skitmore and Marsden (1988) and Singh (1990) used
(4) designing a delivery structure to meet the
the multi-attribute approach, which is a technique
project needs;
applied to measure a degree of objectivity to subjective
(5) allocating responsibilities within the project
areas. Both studies adapted the procurement path
structure; and
decision chart from NEDO (1985) to aid the deci-
(6) establishing a method of appointing for the
sion-making process. Bennett and Grice (1990) have
various participants involved.
undertaken similar work. Furthermore, Skitmore
and Marsden (1988) applied concordance analysis and Essentially, the organization would select the project
Downloaded by [McGill University Library] at 10:57 12 December 2012

discriminant analysis to their theoretical framework. structure and then adapt a contractual arrangement
Concordance analysis is used to measure the consis- to suit. Consequently, such an approach was found to
tency of experts’ ranking for each procurement against a stimulate teamwork in one particular project reported
set predetermined criterion. Discriminant analysis by Love (1996) inasmuch as project goals and objec-
examined data collected under a set of criteria that tives were clearly de® ned. Contrary to this approach,
are characteristics on which the various procurement Hibberd and Basden (1996) suggest that a contractual
methods are expected to differ. Thus, procurement arrangement initially should be selected so as to take
paths could be discriminated against for decision-mak- into consideration how risk will be transferred between
ing purposes. Skitmore and Marsden (1988) found from parties, therefore determining the nature of the
their exploratory work that the multi-attribute approach procurement method so as to ful® l the client’ s objec-
gave similar results to those using discriminant analysis. tives. In essence, Hibberd and Basden (1996) suggest
A cognitive perspective to procurement selection was that risk is the prominent criterion that will determine
proposed by Liu (1994). Organizational behaviour is the selection of a procurement method. Risk alloca-
modelled as an act-to-outcome process. The act-to- tion/avoidance cannot be overlooked; nonetheless,
product and product-to-outcome paths are the project whether or not clients do weight risk as their promi-
realization stage and post-occupancy stage of the nent criterion when procuring a building or other
building procurement process, respectively. Liu (1994) particular criterion will be explored in this paper.
states that organizational behaviour is governed by
organizational goals, and project goals affect the act-
Multi-attribute utility analysis
to-outcome process. A number of moderators such as
ability, task complexity and situational constraints Multi-attribute utility analysis is a methodology that can
affect this goal± performance relationship, thus affecting be used as a tool to measure objectivity in an otherwise
the act-to-product and product-to-outcome processes. subjective area of management (Fellows et al., 1983).
It is suggested by Liu (1994) that, in selecting an Since a procurement system is the overall managerial
appropriate procurement method, the decision maker approach by which a client commissions and obtains a
should take into account the effect of these modera- building, the multi-attribute approach was considered
tors using conjoint analysis (a technique used to model to be the foremost technique appropriate for examining
a decision maker’ s judgement pro® le). the criteria of clients and the preferences of experts’
The procurement module of the `Elsie’ expert system weights for each method in the most objective way. By
computer package (Brandon et al., 1988) provides the indicating the relative utility of each client requirement
recommendations on the most appropriate procure- and procurement method against a numerical scale, it is
ment method via a software program. A series of possible to obtain a set of utility factors. Clients were
questions relating to the timing, quality, design cost invited to give a rating to the above criteria for the latest
parameters and other characteristics of the project is building project they had procured. Quantity surveyors,
posed by the program. On evaluation of the informa- architects, project managers and contracting organiza-
tion, recommendations are given by means of a list tions gave ratings for the above criteria against each pro-
of the most appropriate methods, ranked in order of curement method listed herein. Each procurement
suitability, together with an indication of the extent method and client criterion was scored on a scale of
to which the various methods will satisfy the client 10± 110 to avoid any possible imbalances due to the
requirements. occurrence of zeros (Fellows and Langford, 1980).
Procurement selection 225

Methodology labelled `results’ . The sum total of these for each

procurement option is shown in the last row, together
The research involved an empirical study similar to with the rank order of total. In this example, the best
those of Skitmore and Marsden (1988) and Singh procurement option is novation, with a sum total of
(1990). This involves identifying an appropriate form 84.59, followed by traditional lump sum, with a sum
of procurement path (method) by deriving a set of total of 82.80. The least appropriate procurement
main client requirements and suitable weights option is traditional cost reimbursement, with a sum
according to the project circumstances. This was done total of 37.12.
by means of a postal questionnaire and distributed to
a variety of personnel involved in the procurement
Concordance analysis
process. The weightings obtained from the client
respondents were then juxtaposed with weightings of Numerous classi® cation approaches of client types
`experts’ for each procurement method for a series of have been reported in the construction management
actual completed projects, to derive the theoretically literature. For example, Nahapiet and Nahapiet (1985)
most appropriate form of procurement method for each considered that the needs of clients are in¯ uenced
Downloaded by [McGill University Library] at 10:57 12 December 2012

project. Then this was compared with the procurement by two important characteristics: whether they are
method actually used for the project. `primary or secondary constructors’ and `their level of
project experience’ . Masterman and Gameson (1994)
suggest that clients should be classi® ed according to
The procedure for weighting criteria
their construction experience and whether they are
Different clients and different project circumstances primary or secondary constructors. When these two
demand different criteria weights. For example, if, for characteristics are combined, the following client types
one project, the cost is the most important aspect, then are created: secondary inexperienced; secondary expe-
we would weight the `cost’ criterion higher than the rienced; primary inexperienced and primary experi-
other criteria. For another project where the speed of enced (Table 1).
construction is the most important, we would weight Client experience will continually change. Every pro-
the `speed’ criterion higher than the other criteria. ject is different. Therefore it may be considered that
The procedure adopted for obtaining client priority there is no such thing as an `experienced’ client in the
weightings for each criterion follows Singh (1990). strictest sense, although clients may acquire a degree of
This involves the following steps. (1) The client knowledge and understanding of the environment
weights the relative importance of each criterion (i.e. within which the project is being procured. Further-
speed, certainty, ¯ exibility) on a scale of 1± 20. This more, it is critical that clients understand where their
relative importance score is termed a priority rating. boundaries of construction expertise lay (Masterman
(2) Rationalized priority ratings are then calculated (by and Gameson, 1994). The study described herein
dividing each priority rating by the sum of all the
ratings). The sum of the rationalized priority ratings Table 1 Masterman and Gameson’ s (1994) client type
then will always be equal to 1. (3) Each rationalized classi® cation
priority rating is taken in turn and multiplied by Classi® cation De® nition
a utility factor representing the extent to which a
procurement method satis® es a criterion. The utility Primary Clients such as property developers, whose
factors connect each criterion to each procurement main business and
primary income derive from constructing
method in a consistent way, irrespective of the project.
Thus, the traditional procurement method, which is
Secondary Clients for whom expenditure on construct-
known to be fairly slow, is given a fairly low utility ing buildings is a small
factor score. The construction management procure- percentage of their turnover, and from
ment method, on the other hand, which is known to whom buildings are necessary in order
be fairly fast, is given a fairly high utility factor score. to undertake a speci® c business activity,
(4) The rationalized priority rating± utility factor prod- such as manufacturing.
ucts are added for each procurement method and the Experienced Recent and relevant experience of con-
resulting total ranked in descending order. The most structing certain types of building,
appropriate procurement method is taken to be the one with established access to construction
with the highest total. expertise either in-house or externally.
Inexperienced No recent and relevant experience of
An example is the procurement path decision chart
constructing buildings with no established
shown in Figure 2. In the chart the rationalized priority
access to construction expertise.
rating± utility factor products are entered in the column
Downloaded by [McGill University Library] at 10:57 12 December 2012

Figure 2 Example of a procurement path decision chart

Love et al.
Procurement selection 227

implemented a classi® cation whereby clients were clas- imately 25% were property and development compa-
si® ed according to their knowledge of the construction nies, 25% investors, 30% occupiers, and 20% local and
process, and market factors. Consequently, the follow- central government authorities (including 7% from
ing classi® cation of clients’ market and technical knowl- local authorities). These results indicate an even spread
edge of the construction environment was used: of clients. Once the deadline date for the return of the
questionnaires had passed, those clients who had
(1) good knowledge of both technical and market
replied to the questionnaires were contacted and inter-
viewed via telephone. All 41 clients cooperated in the
(2) good technical knowledge but limited or no
follow-up interview.
market knowledge;
(3) limited or no technical knowledge but a ® rm
understanding of market factors; and Consultants
(4) limited or no knowledge of both technical and
Questionnaires were mailed to 100 selected consultants
market factors.
throughout Australia. Consultants were given over one
Ideally, we would like the clients’ responses to be month to reply to the questionnaire. Only 10 ques-
Downloaded by [McGill University Library] at 10:57 12 December 2012

homogeneous so the data for the utility factor scores tionnaires were returned within the time allocated.
can be pooled for all of these classi® cation groups. Each consultant was then telephoned to establish
Concordance analysis is used to test this, and involves why they had not returned the questionnaire. As a
calculating the coef® cient of concordance (Kendall and result of the telephone calls, a further 10 question-
Babington-Smith, 1939) to see if the rankings provided naires were returned. Questionnaires also were mailed
by the client classes in weighting each criterion are to another 50 consultants throughout Australia who
in suf® cient agreement. This statistical technique were given one month to return the questionnaire.
measures the rank correlation for a number of rank- Fifteen questionnaires were returned, bringing the total
ings. The measure for the coef® cient of concordance sample size to 35.
is de® ned by:
W = 12Sw /m2(n3 ± n)
where Sw equals the sum of the squares of the devia-
tions of the total of the ranks assigned to each indi- Data from the questionnaires were extracted to derive
vidual from m(n)/2. The quantity m(n + 1)/2 is the weightings of utility factors. These weightings then
average value of the totals of the ranks, and hence Sw were examined to determine whether or not the
is the sum of squares of deviations from the mean. respondents gave similar weights for the same criterion
W varies from 0 to 1; 0 represents no community of for differing project types.
preference, and 1 represents perfect agreement. Using
Kendall and Babington-Smith’ s de® nition of the coef-
® cient of concordance in this study, it was applied in
the following way: m is the number of observers, n is The most common procurement method used by the
the number of procurement categories, and Sw is the client respondents is the traditional lump sum and
sum of the ranks for each procurement method, and documentation (56%), with novation the next most
the deviation of each sum from the average is then popular system (18%), and the management system
calculated. of design, manage and construct the least used (3%).
74% of clients procured their development less than
one year previously. Clients were classi® ed into
Data collection either investors, property and development companies,
local and central government authority or occupiers
Clients (Turner, 1990).
Clients weighted each criterion using the scale indi-
Questionnaires were mailed to 100 selected clients cated in the questionnaire. Each client type was classi-
throughout Australia (except for the Northern ® ed in terms of their experience of market and technical
Territory). Clients were given over one month to reply knowledge of the construction industry (Table 2). The
to the questionnaire. Initially only 20 clients returned time period as to when they had completed their last
their questionnaires, therefore a further 50 question- building project was used as the basis for verifying
naires were mailed to clients. These clients were given clients and their perceived satisfaction or dissatisfaction
two weeks to answer the questionnaire, resulting in a with the form of procurement method actually imple-
total of 41 being received eventually, of which approx- mented.
228 Love et al.

Table 2 Summary of client classi® cation

Clienta Market and technical knowledge Building type %

A Good knowledge of both technical and market factors Commercial 80
Good technical knowledge but limited or no market knowledge Residential 20

B Good knowledge of both technical and market factors Residential 30

Commercial 50
Recreational 20

C Good knowledge of both technical and market factors Other (airports) 18

Educational 9
Limited or no technical knowledge but a ® rm understanding of market factors Industrial 9
Limited or no knowledge of both technical and market factors Commercial 9
Admin. and civic 9
Limited or no technical knowledge but a ® rm understanding of market factors Educational 18
Downloaded by [McGill University Library] at 10:57 12 December 2012

Limited or no knowledge of both technical and market factors Educational 18

Hospital 9

D Good knowledge of both technical and market factors Hospital 25

Admin. and civic 12.5
Residential 12.5
Limited or no technical knowledge but a ® rm understanding of market factors Educational 25
Limited or no knowledge of both technical and market factors Recreational 12.5
Hospital 12.5
Key: A, investors; B, property and development companies; C, occupiers; and D, local and central government authorities

The classi® cations of building types which clients Table 3 Mean weights and ranks of client typesa
procured are as follows:
Criteria A B C D
(1) residential;
Weight Rank Weight Rank Weight Rank Weight Rank
(2) commercial;
(3) recreational; Speed 13 6 20 1 14 5 17 4
(4) administration and civic; Certainty 16 4 18 3 15 5 19 2
(5) industrial; Flexibility 15 5 12 7 9 8 14 6
(6) hospital; Quality 13 6 15 6 17 4 18 4
(7) educational; and Complexity 8 8 9 8 10 7 11 8
Responsibility 18 3 17 4 16 4 13 7
(8) other.
Each client’ s weights were ranked in order of prefer- and dispute 8 8 9 8 17 3 14 6
ence. The mean weights of each client type were calcu- Price competition 17 4 16 5 19 2 20 1
lated along with the corresponding mean rank (Table Risk allocation/
3), and the coef® cient of concordance W was calcu- avoidance 18 2 17 4 12 7 13 7
lated for each client type (Table 4). All the results are a
Key: A, investors; B, property and development companies; C,
below the critical value of 0.70, and therefore indicate occupiers; and D, local and central government authorities.
that there is insuf® cient consistency with the weight-
ings for the utility factors. As a result, it was concluded Table 4 Coef® cients of concordance for clients
that different clients have different needs.
The clients indicated their satisfaction with the Client type Coef® cient of
procurement method adopted, as shown in Table 5. concordance
It is indicated that 70% of clients who were satis® ed
with the procurement of their buildings used a design Investors 0.57
and build procurement system. Furthermore, it was Property and development companies 0.27
found that these clients utilized an independent project Occupiers 0.25
manager to act as their project representative and Local and central government authorities 0.64
Procurement selection 229

Table 5 Percentage of clients satis® ed with their procure- technical speci® cation and quality; (c) lack of feed-
ment method back from participants to the project’ s performance;
Client type Novation Design Contractors Traditional (d) lack of involvement throughout the project; (e)
and design and lump poor coordination and communication between partic-
manage build sum ipants; (f) con¯ icting advice from consultants; and (g)
no project manager to act as an integrating device
Investors ± ± ± 10
between participants.
Property and
Evidently, from the results obtained, there are partic-
companies ± 15 15 ±
ular factors that contribute to client satisfaction and
Occupiers 35 5 ± ± these should be considered prior to the selection of a
Local and central procurement method.
government A study by Walker (1994b) of the Australian con-
authorities ± ± ± 20 struction industry found that experienced or sophisti-
cated clients are more likely to achieve a successful
project outcome. From this research, a key factor
Downloaded by [McGill University Library] at 10:57 12 December 2012

contributing to project success was also the client’ s

ability to recognize the role and function of an inde-
pendent project management organization that acted
principal adviser. The clients’ responses indicated as a focal point for project participants.
the following fundamental factors contributed to the Masterman and Gameson (1994, pp. 81± 4) have
clients’ satisfaction: (i) completion of the project on identi® ed the desire of the client to be involved and
time and to budget; (ii) completion to the desired tech- informed about the project, referring to studies of
nical speci® cation and quality; (iii) teamwork and client needs by Bennett and Flanagan (1983) and
commitment from all participants; (iv) ability of partic- Hewitt (1985), which found that clients needed active
ipants to understand the goals and objectives of the involvement and to be kept informed. The lack of
project; (v) effective communication both formally and involvement by the client was identi® ed as a factor con-
informally between participants; and (vi) an indepen- tributing to their dissatisfaction. This factor was com-
dent project manager. mon amongst clients who classi® ed themselves as
Those clients who were dissatis® ed with their having a good knowledge of both technical and market
procurement method were those who had implemented factors and those with limited or no knowledge of the
construction management and traditional lump sum industry. Masterman (1994), in a detailed study of the
methods (Table 6). Completion of the building within basis upon which clients select a procurement method,
a stipulated time period was a fundamental reason found that the need of the client to be involved and
why clients implemented a construction management informed of the ® nal cost and certainty of completion
method. This led to other criteria being neglected, was amongst the ® rst of their priorities when selecting
thus other priorities of the client were not evaluated a procurement route.
in a holistic manner as they had not employed the In this research, it was found that experienced clients
skills of a project manager. The following fundamen- have different needs from inexperienced clients in their
tal factors were identi® ed as contributing to clients’ procurement objectives. The factors identi® ed by
dissatisfaction (a) project not completed on time nor Masterman (1994) were, however, considered in hind-
to budget; (b) project not completed to the desired sight by the client when they were dissatis® ed with
the procurement process. Gameson (1992, pp. 203± 7)
has shown that construction professionals tend to take
a dominant and diagnostic role in the relationship
Table 6 Percentage of clients dissatis® ed with their with inexperienced clients, but that role becomes less
procurement method supportive and less in¯ uential when dealing with expe-
rienced clients. Consultants may agree on the funda-
Client type Construction Traditional
mental objectives of time, cost and quality, but place
management lump sum
emphasis on the performance standards that affect
Investors 10 28 their own expertise. This was found to be the case with
Property and development dissatis® ed clients who had implemented a procure-
companies 5 28 ment system on the advice of their consultants rather
Occupiers 10 ±
than an independent project manager who could take
Local and central government
a holistic approach of the clients’ strategic project
authorities ± 19
230 Love et al.

Consultants The mean values of the consultants’ utility weight-

ings for each criterion against each procurement
Thirty-® ve consultants completed and returned the
method are shown in Table 7. The results indicate
questionnaires. Of these, 14% were architects, 57%
that method A (traditional lump sum) provides the
project managers and 29% quantity surveyors. The
best quality (mean weighting 100.00) and best price
most popular method of procurement for the whole
competition (mean weighting 94.50); method E
sample of respondents was found to be traditional
(turnkey and package deals) is the most certain (mean
lump sum (42%), followed by novation (34%),
weighting 100.00), best for risk allocation/avoidance
contractor design and build (16%), and turnkey and
(mean weighting 109.70) and best for responsi-
package deal (8%). All consultants thought that the bility (mean weighting 95.60); method F (novation) is
reasons for the popular forms were: (1) clients want the best for avoiding arbitration and disputes (mean
to reduce the amount of risk they are willing to take weighting 95.60); and method G (construction man-
due to the prevailing economic climate; (2) clients’ agement) is the speediest (mean weighting 90.50),
main priority is cost and certainty in times of reces- most ¯ exible (mean weighting 95.60) and best for
sion; and (3) clients require lump sum before construc- complexity (mean weighting 105.0). The smallest coef-
tion commences.
Downloaded by [McGill University Library] at 10:57 12 December 2012

® cient of concordance was 0.61 (for arbitration and

The least popular method of procurement for disputes), and this was taken to indicate the existence of
these respondents was found to be: (a) manage- a reasonable consensus by the consultants on the value
ment method ± design, manage and construct; (b) of weightings (Table 8).
management method ± management contracting; Consultants (e.g. architects, quantity surveyors)
(c) traditional method ± cost reimbursement; and (d) prefer to use a traditional lump sum method instead
traditional method ± provisional quantities. All consul- of non-traditional methods. Essentially, the traditional
tants considered systems (b), (c) and (d) to be not lump sum method requires a project to be fully
popular within the marketplace, with 90% stating that designed and documented prior to tendering. This may
method (a) was unpopular. require greater fees than non-traditional methods,
The architects and quantity surveyors subsamples therefore increasing the potential for pro® ts to be made
principally weighted the traditional method of pro- by consultants. Conversely, if non-traditional methods
curement with the higher utility preferences (except are implemented by the client, the services of consul-
in the case of cost reimbursement), whereas the tants become restricted, primarily because of the intro-
project managers subsample tended to show no pref- duction of the contractor. For example, under the
erence towards any particular system; their approach traditional lump sum method, architects predomi-
to weighting each parameter against each procure- nantly act as the lead consultant, primarily being
ment method tended to be impartial (i.e. show no responsible for developing the brief, managing and
favouritism). There was no doubt that architects’ and controlling the design process, and acting as a super-
quantity surveyors’ priority weightings favoured tradi- visor for the client during construction. Therefore this
tional lump sum and traditional lump sum with provi- may require more fees than the restricted role that
sional quantities. is undertaken by an architect when a non-traditional

Table 7 Mean utility factors of criteria for procurement methods

Criteria Procurement methodsa

Speed 52.5 56.5 45.6 76.2 79.6 83.5 90.5 88.6 81.5
Certainty 88.5 80.6 29.1 90.3 100.0 85.6 55.6 50.2 53.8
Flexibility 75.6 86.3 65.2 59.6 45.0 73.8 95.6 94.8 85.2
Quality 100.0 95.6 58.3 60.5 45.5 85.2 73.6 71.2 84.5
Complexity 80.6 78.5 55.0 75.6 50.0 95.3 105.0 100.0 85.5
Risk allocation/
avoidance 80.0 70.0 10.0 96.8 109.7 92.5 45.0 40.0 50.0
Responsibility 88.6 75.2 20.0 92.5 95.6 90.5 36.0 35.8 40.0
disputes 75.3 65.3 10.0 70.8 83.5 95.6 58.3 55.2 57.6
Price competition 94.5 76.7 44.7 42.0 40.0 62.5 90.0 90.0 80.0
Key: A, traditional single lump sum; B, traditional provisional quantities; C, traditional cost reimbursement; D, contractors design and
build; E, turnkey and package deals; F, novation; G, construction management; H, management contracting; and I, design and manage.
Procurement selection 231

Table 8 Coef® cients of concordance for consultants and construct method, tend to be consistent through-
out all the charts (ranks 3± 5, but on occasion exchange
Selection criterion Coef® cient of concordance
(W) places). Construction management, management con-
tracting, and design and construct occupied ranks 6, 7
Speed 0.73 and 8, with management contracting predominantly
Certainty 0.85 being ranked eighth and the positions of the remaining
Flexibility 0.99
management methods exchanging ranking positions.
Price competition 0.65
The traditional cost reimbursement form was ranked
Risk allocation/avoidance 0.96
Responsibility 0.86 ninth for all the charts.
Quality 0.70
Arbitration and disputes 0.61
Complexity 0.94 Conclusion

It is shown that similar clients do not in general have

method is adopted, that is, simply acting as a designer. similar needs in their procurement objectives. This may
Downloaded by [McGill University Library] at 10:57 12 December 2012

For the quantity surveyor, preparation of bills of quan- of course be due to the different nature of their indi-
tities has traditionally been their primary source of vidual projects; whether the same client has the same
income, and any change in the way in which projects needs for different projects is not examined here. There
are delivered may hinder their incoming ¯ ow of monies is a consensus, however, that the criteria proposed,
and pro® t. Clearly, consultants will recommend a and their weights, are themselves appropriate for each
procurement system that meets their needs, and not procurement method. It was surprising to ® nd that the
necessarily the immediate needs of the client. However, application of these weights in the procurement path
clients appear to be placing greater emphasis on single- decision charts resulted in the same procurement deci-
point responsibility, risk avoidance and ¯ exibility, sion (novation) to be the highest ranked for all the test
rather than on quality and cost, which are considered projects used, and suggests that a further study is
to be the fundamental attributes of the traditional lump needed to validate and understand this, perhaps by
sum method. Design and build systems appear to in-depth interview. It should be noted, however, that
accommodate these increasing needs (Bresnen et al., the two most common procurement methods used
1987), but whether or not consultants will recommend by the client respondents are the traditional and nova-
the use of design and construct is perhaps question- tion, and the procurement path decision chart found
able. Therefore, based on the ® ndings presented above these also to be the most appropriate, albeit in reverse
and to eliminate bias, we suggest that an independent order.
project manager selects the procurement method by The low rankings for construction management,
systematically weighting client needs in accordance management contracting and cost reimbursement may
with their preferences. be due to the intrinsic uncertainty involved in these
methods. NEDO (1985) suggests that the management
methods system offers price certainty, although, at the
Procurement path decision chart
time of contract, the exact nature and detail of the
A procurement path decision chart (Skitmore and project generally are not established. In our view,
Marsden, 1988) was produced for each client respon- management methods are derived from prime cost
dent using the mean utility values of the consultants’ contracts and are thus lacking in price certainty.
weights from Table 6 juxtaposed with the clients’ The device of a guaranteed maximum price some-
criteria weightings. An example is shown in Figure 2. times is offered, but it is possible to obtain price
Each procurement method was ranked, with the certainty only if the maximum being guaranteed is high
highest result being ranked 1. Method F in this enough, in effect to contain a target ® gure that includes
example, with a total weighting of 84.59, represents suf® cient contingency. A maximum guaranteed price
the `appropriate’ form of procurement method. A total concept is not often possible to obtain before the time
of 41 charts were produced in this way, one for each when a construction contract needs to be signed.
client respondent, and in every case the appropriate The cost reimbursement form is a system whereby the
form of procurement system is the design and construct contractor is paid the actual or prime cost for an inde-
novation form, with the traditional lump sum and terminate amount of work, and in addition an agreed
documentation form ranked as the second choice, irre- fee is paid to cover management, overheads and pro® t.
spective of the type of client or building involved. It is possible that this form is not favoured as the resul-
The positions of ranks for the traditional methods tant ® nal cost to the client is dif® cult to determine.
with provisional quantities and the remaining design As with construction management and management
232 Love et al.

contracting, fees are paid on the actual cost of the work Gameson, R.N. (1992) An investigation into the interaction
undertaken. Moreover, Barclay (1994) found from his between potential building clients and construction profes-
studies that the design, manage and construct form has sionals, unpublished Ph.D. thesis, University of Reading,
not been used that extensively within Australia, hence UK.
Graves, F. (1978) Construction for an Industrial Recovery,
the lower weights and the low overall aggregate rank.
National Economic Development Of® ce (NEDO),
A simple set of criteria has been identi® ed as being
Steering Group on Industrial Building and Infrastructure,
generally adequate and suf® cient for procurement HMSO, London.
selection, and there is a reasonable consensus on the Hamilton, I.W. (1987) Developing expert systems for
appropriate weightings for each path. Moreover, the management applications, in Building Cost Modelling and
current trend in the literature urging greater involve- Computers, P.S. Brandon (ed.), E & FN Spon, London,
ment and interaction between client and consultants pp. 441± 51.
for a more effective procurement process is supported Hewitt, R.A. (1985) The procurement of buildings:
by the results of this study. The juxtaposition of design proposals to improve the performance of the industry,
and construct options with a project management orga- unpublished project report, College of Estate Manage-
nization is seen as a way of generally satisfying clients’ ment, Reading, UK.
Hibberd, P. and Basden, A. (1996) The relationship between
Downloaded by [McGill University Library] at 10:57 12 December 2012

needs. For inexperienced clients in particular, this

procurement and contractual arrangements, in Proceedings
points to the use of a project management organiza-
of CIB-W92 Procurement Systems Symposium, North Meets
tion that acts as a control mechanism for the client South, Developing Ideas, Durban, South Africa, 14± 17
and participants. January, pp. 639± 46.
Ireland, V. (1982) Variables in building projects and the
achievement of objectives, in Proceedings of the P.M.F.
References National Conference, Sydney, Australia, September, pp.
39± 49.
Barclay, J. (1994) A comparison between traditional and Ireland, V. (1985) The role of managerial actions in cost,
non-traditional forms of contracting for the procurement time and quality performance of high rise commercial
of building projects, B.Sc. thesis, School of Construction building projects, Construction Management and Economics,
Management, Queensland University of Technology. 3(1), 59± 87.
Bennett, J. (1986) Construction Management and the Chartered Kendall, M.G. and Babington-Smith, B. (1939) The
Quantity Surveyor, Surveyor Publications, London. problem of m ranking’ s, Annals of Mathematical Statistics,
Bennett, J. and Flanagan, R. (1983) For the good of the 10, 275± 87.
client, Building, 1 April, 26± 7. Liu, A.N.N. (1994) From act to outcome ± a cognitive model
Bennett, J. and Grice, A. (1990) Procurement systems for of construction procurement, in Proceedings of CIB W-92
buildings, in Quantity Surveying Techniques: New Directions, International Procurement Symposium, East Meets West,
P.S. Brandon (ed.), BSP Professional Books, Oxford. Department of Surveying, University of Hong Kong, 4± 7
Blackmore, C. (1990) The Clients Tale, RIBA Publications, December, pp. 169± 78.
London. Love, P.E.D. (1996) Fast building: an Australian perspec-
Brandon, P.S., Basden, A., and Hamilton, I.W. (1988) tive, in Proceedings of CIB-W92 Procurement Systems
Expert System: The Strategic Planning of Construction Projects, Symposium, North Meets South, Developing Ideas, Durban,
Royal Institution of Chartered Surveyors and University South Africa, 14± 17 January, pp. 329± 43.
of Salford. Love, P.E.D., and Mohamed, S. (1996) Project management:
Bresnen, M.J. and Haslam, C.O. (1991) Construction the key to procuring fast buildings, Asia Paci® c Building
industry clients: a survey of their attributes and project and Construction Management Journal, 2(1), 1± 5.
management practices, Construction Management and Masterman, J.W.E. (1992) An Introduction into Building
Economics, 9(4), 327± 42. Procurement System, E & FN Spon, London.
Bresnen, M.J., Haslam, C.O., Beardsworth, A.D., Bryman, Masterman, J.W.E. (1994) A study of the basis upon
A.E. and Kiel, E.T. (1987) Performance on Site and The which clients of the construction industry choose their
Building Client, Occasional Paper No. 42, Chartered building procurement systems, unpublished Ph.D. thesis,
Institute of Building, Englemere, Kings Ride, Ascot. University of Manchester, Institute of Science and
Fellows, R.F. (1993) Contracts for Refurbishment, School of Technology.
Architecture and Building Engineering, University of Bath. Masterman, J.W.E., and Gameson, R.N. (1994) Client char-
Fellows, R.F. and Langford, D.A. (1980) Decision theory acteristics and needs in relation to their selection of
and tendering, Building Technology and Management, building procurement systems, in Proceedings of CIB W-92
October, 36± 9. International Procurement Symposium, East Meets West,
Fellows, R.F., Langford, D.A., Newcombe, R., and Urry, S. Department of Surveying, University of Hong Kong, 4± 7
(1983) Construction Management in Practice, Longman, December, pp. 79± 88.
London. Nahapiet, H. and Nahapiet, J. (1985) A comparison of
Franks, J. (1990) Building Procurement Systems, Chartered contractual arrangements for building projects, Construction
Institute of Building, Englemere, Kings Ride, Ascot. Management and Economics, 3(3), 217± 31.
Procurement selection 233

NEDO (1975) The Public Client and the Construction Skitmore, R.M. and Marsden, D.E. (1988) Which procure-
Industries, National Economic Development Of® ce, ment system? Towards a universal procurement selection
HMSO, London. technique, Construction Management and Economics, 6(1),
NEDO (1983) Faster Building for Industry, National 71± 89.
Economic Development Of® ce, HMSO, London. Turner, A. (1990) Building Procurement, Macmillan, London.
NEDO (1985) Think About Building, National Economic Walker, A. (1989) Project Management in Construction, BSP,
Development Of® ce, HMSO, London. London.
Perry, J.G. (1985) The development of contract strategies Walker, D.H.T. (1994a) An investigation into factors
for construction projects, Ph.D. thesis, University of that determine construction time performance, unpub-
Manchester. lished Ph.D. thesis, Royal Melbourne Institute of
Shockley-Zalaback, P. (1991) Fundamentals of Organisational Technology.
Communication. Longman, New York, pp. 7± 8. Walker, D.H.T. (1994b) Procurement systems and con-
Singh, S. (1990) Selection of appropriate project delivery struction time performance, in Proceedings of CIB W-92
system for construction projects, in Proceedings of CIB W- International Procurement Symposium, East Meets West,
90 International Symposium on Building Economics and Department of Surveying, University of Hong Kong, 4± 7
Construction Management, Sydney, Australia, pp. 469± 80. December, p. 343.
Downloaded by [McGill University Library] at 10:57 12 December 2012